{
 "cells": [
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "# Syntactic similarity metrics",
   "id": "eb9fb4a1a68e633"
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "AMLGym provides several measures for evaluation an action model w.r.t. a reference model or simulator.",
   "id": "1ac309b575ed4764"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": "!pip install amlgym",
   "id": "3a2c1c2a35c2e0f2"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": [
    "from amlgym.metrics import print_metrics\n",
    "\n",
    "print_metrics()"
   ],
   "id": "5074d7d2f20b291f"
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "Learn a model to be evaluated:",
   "id": "243b51efc337c73b"
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "initial_id",
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from amlgym.benchmarks import get_domain_path, get_trajectories_path\n",
    "from amlgym.algorithms import get_algorithm\n",
    "from amlgym.util.util import empty_domain\n",
    "\n",
    "# Get learning algorithm\n",
    "learner = get_algorithm('rosame')\n",
    "\n",
    "# Get an input domain with no preconditions and effects\n",
    "domain_ref_path = get_domain_path('blocksworld')\n",
    "domain_empty_path = empty_domain(domain_ref_path)\n",
    "\n",
    "# Get a set of trajectories to learn from\n",
    "traj_paths = get_trajectories_path('blocksworld')\n",
    "\n",
    "# Learn a domain model\n",
    "model = learner.learn(domain_empty_path, traj_paths)"
   ]
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "Print the learned model",
   "id": "93a7b7785b9cd873"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": "print(model)",
   "id": "a54bbcad9ad6fe7f"
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "Save the learned model to a file:",
   "id": "e57ba7ed0e528f13"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": [
    "domain_eval_path = 'domain_learned.pddl'\n",
    "with open(domain_eval_path, 'w') as f:\n",
    "    f.write(model)"
   ],
   "id": "180a5e01e627481f"
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "Syntactic similarity can be measured by means of precision and recall:",
   "id": "79155f5492207006"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": [
    "from amlgym.metrics import syntactic_precision\n",
    "precision = syntactic_precision(domain_eval_path, domain_ref_path)\n",
    "print(precision)"
   ],
   "id": "fd36b0d1fc890673"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "outputs": [],
   "execution_count": null,
   "source": [
    "from amlgym.metrics import syntactic_recall\n",
    "recall = syntactic_recall(domain_eval_path, domain_ref_path)\n",
    "print(recall)"
   ],
   "id": "17a95e4b15c731e8"
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}